Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Recurso digital |
| Lingua: | |
| Pubblicazione: |
Zenodo
2025
|
| Accesso online: | https://doi.org/10.5281/zenodo.17934434 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
Sommario:
- <p><span lang="EN-US">Human cognition displays a remarkable balance between stability and flexibility: we maintain a continuous sense of self while rapidly shifting among tasks, intentions, and contexts. Yet dominant theories—such as Global Workspace Theory, Predictive Coding, and control-network models—primarily describe how information is processed once it becomes relevant, rather than how the cognitive system determines what it is currently organized around. This paper introduces the Cognitive Anchor System, a structural framework proposing three organizing anchors that define the moment-to-moment architecture of cognition. The self anchor provides the first-person perspective and the experiential continuity that grounds all cognitive activity. The task anchor specifies the present goal and structures attention, contextual information, and resource allocation. Surrounding these is an anchor pool consisting of latent intentions, pending goals, unresolved concerns, and background plans that dynamically compete for elevation to the task anchor. Through activation, inhibition, competition, and reconfiguration, these anchors shape task selection, task switching, mind wandering, procrastination, decisional conflict, insight, and the maintenance of long-term goals. By comparing this model with major cognitive theories, we show that anchor structures supply a missing organizational layer that determines which contents gain access to conscious processing and behavioral control. The framework generates testable predictions for behavioral experiments, neurocognitive dynamics, and future cognitive architecture design. The Cognitive Anchor System offers a unifying structural principle that links diverse cognitive phenomena within a coherent and empirically tractable model.</span></p>